How to Grep a Word Exactly

Grepping for exact words with UNIX

Word boundary matching is an extension to the standard POSIX grep utility. It might be available or not. If you want to search for words portably, I suggest you look into perl instead, where you would use

perl -ne 'print if /\baaa\b/' $EAT_Setup_BJ3/Log.txt

How to grep for the whole word

You want the -w option to specify that it's the end of a word.

find . | xargs grep -sw 's:text'

Display exact matches only with grep

You need a more specific expression. Try grep " OK$" or grep "[0-9]* OK". You want to choose a pattern that matches what you want, but won't match what you don't want. That pattern will depend upon what your whole file contents might look like.

You can also do: grep -w "OK" which will only match a whole word "OK", such as "1 OK" but won't match "1OK" or "OKFINE".

$ cat test.txt | grep -w "OK"
1 OK
2 OK
4 OK

Match exact word using grep

Try this:

grep -E "^test1" file

This states that anything that starts with the word test1 as the first line. Now this does not work if you need to fin this in the middle of line but you werent very specific on this. At least in the example given the ^ will work.

How to grep for the exact word if the string has got dot in it

You need to escape the . (period) since by default it matches against any character, and specify -w to match a specific word e.g.

grep -w -l "BML\.I" *

Note there are two levels of escaping in the above. The quotes ensure that the shell passes BML\.I to grep. The \ then escapes the period for grep. If you omit the quotes, then the shell interprets the \ as an escape for the period (and would simply pass the unescaped period to grep)

grep for exact match

If you're using GNU grep then you can use the -w or --word-regexp flag, or wrap the regex with word boundaries:

grep '\bH\b'

However, this will be non-portable, since POSIX grep lacks the -w flag and neither basic nor extended regular expressions support word boundaries. So if you wish for the command to be portable then you'll have to do something like

grep ' H '

or otherwise use a more powerful language, like AWK:

awk '$2=="H" { print }'

Exact Match of Word using grep

how about an awk solution?

awk -F'|' '$1 == "CHICAGO"{print $2}' file

to only print one output, exit once you have a match, i.e.

awk -F'|' '$1 == "CHICAGO"{print $2; exit}' file

Making that more generic, you can pass in a variable, i.e.

awk -v trgt="CHICAGO" -F'|' '{targ="^" trgt " *$"; if ( $1 ~ targ ) {print $2}}' file

The " *$" regex limits the match to zero or more trailing spaces without any extra chars at the end of the target string. So this will meet your criteria to match skip matching CHICAGO PIT|BULL.

AND this can be further reduced to

awk -v trgt="CHICAGO" -F'|' '{ if ( $1 ~ "^" trgt " *$" ) {print $2}}' file

constructing the regex "in-place" in with the comparison.

So you could use more verbose variable names to "describe" how the regex is being constructed from the input and the regex "wrappers" (as in the 3rd example) OR, you can just combine the input variable with the regex syntax in place. That is just a matter of taste or documentation conventions.

You might want to include a comment to explain you are constructing a regex test that would look like the $1 ~ /^CHICAGO *$/.

IHTH

grep,how to search for exact pattern?

If by "exact pattern" you mean a complete line with only qtrain in it,
then use the -x flag:

grep -nx qtrain *.m

This will only match lines that contain exactly "qtrain" and nothing else.

---

If by "exact pattern" you mean to match "qtrain" but not "blahqtrain" or "qtrainblah", then you can use -w to match whole words:

grep -nw qtrain *.m

This will match only this line in your input:

SKSC.m:195:model.qtrain = model.qtrainExtra(:,k-1);

Btw, here's another equivalent way using regular expressions:

grep -n '\<qtrain\>' *.m

From man grep:

The symbols \< and \> respectively match the empty string at the
beginning and end of a word.

---

Related Topics